Hypoglycaemia in non-diabetic pregnancy

Abstract

Hypoglycaemia in non-diabetic pregnancy is rare, the majority of reported cases being due to insulinoma, acute fatty liver of pregnancy, malaria and inborn errors of metabolism. A case of hypoglycaemia in a woman with previous laparoscopic sleeve gastrectomy, and hypothalamic-pituitary-adrenal axis insufficiency in the setting of opioid dependence is presented. The timing of low interstitial glucose levels was atypical for late dumping syndrome following bariatric surgery, and a change in the woman’s glucocorticoid replacement resulted in resolution of hypoglycaemic symptoms. The incidence of opioid dependence in pregnancy is increasing rapidly. Health professionals should be aware of the possibility of opioids causing hypothalamic-pituitary-adrenal axis insufficiency, and the additional mechanisms by which opioids may cause hypoglycaemia.

Introduction

Hypoglycaemia in non-diabetic pregnancy is rare, the most common cause globally being falciparum malaria, which is associated with hypoglycaemia in up to 36% pregnant women. Fourteen cases of insulinoma have been described during pregnancy, 10 of these presenting in first trimester. Additionally, 11 cases of insulinoma have been described in the early postpartum period. Venous glucose less than 4 mmol/L has been described in 61–71% of cases of acute fatty liver of pregnancy. Single case reports of hypoglycaemia in pregnancy have been described with non-islet cell tumour hypoglycaemia and drug-induced insulin autoimmune syndrome. The incidence of hypoglycaemia during pregnancy in women with previous bariatric surgery is not known, with only three case reports described in the literature. Similarly, the incidence of hypoglycaemia in pregnant women with hypothalamic-pituitary-adrenal axis insufficiency (HPAi) is not known. Maternal hypoglycaemia is associated with risk of maternal injury, as well as fetal growth restriction. A case of recurrent hypoglycaemia in a woman with prior bariatric surgery and opioid-associated HPAi is described.

Case

A telephone consultation was performed with a 31-year-old non-diabetic woman, G3 P0, at 22 weeks of gestation regarding the recent onset of symptomatic hypoglycaemia, with capillary blood glucose as low as 1.8 mmol/L. Symptoms typically occurred 1–3 h after meals, resolving with ingestion of carbohydrate.

The woman’s history included chronic pain, opioid dependence, and laparoscopic sleeve gastrectomy (LSG) three years earlier. Twelve months prior to conception HPAi was diagnosed; with no other abnormalities of pituitary function identified and pituitary morphology was normal on imaging. The most likely diagnosis was opioid-induced HPAi. Medications for the woman were hydrocortisone 12 mg in the morning and 8 mg in the afternoon (0.24 mg/kg bodyweight), buprenorphine, quetiapine, venlafaxine and diazepam. She denied use of alcohol, insulin, sulphonylureas, complementary or herbal therapies, and reported adherence with prescribed medications.

Diagnostic possibilities considered included late dumping syndrome post-LSG, inadequate hydrocortisone replacement (dose and/or frequency), opioid-induced hypoglycaemia, nesidioblastosis, insulinoma and factitious hypoglycaemia. Due to her rural location and social circumstances, hospital admission to document venous hypoglycaemia with paired hormone levels, and a cortisol day curve to assess adequacy of glucocorticoid replacement was not practicable. A morning serum cortisol was 185 nmol/L (normal 2nd trimester > 450 nmol/L) with ACTH 2.6 pmol/L (5–13 pmol/L) consistent with HPAi. Hepatic function was normal. A low-carbohydrate diet was prescribed as late dumping was thought to be the most likely diagnosis. Flash glucose monitoring, however, revealed frequent episodes of low interstitial fluid glucose both fasting and pre-prandially, atypical for dumping syndrome, and more suggestive of inadequate glucocorticoid replacement, insulinoma or opioid-induced hypoglycaemia (Figure 1). The woman felt her medication adherence would be poor with hydrocortisone three times daily, thus her glucocorticoid replacement was changed to prednisone 10 mg in the morning with complete resolution of symptoms. The pregnancy proceeded uneventfully until 34 weeks and 3 days of gestation when the mother presented with preterm premature rupture of membranes and premature labour, delivering a live female birthweight 2130 g. The neonatal course was complicated by respiratory distress requiring non-invasive ventilation for 24 h. Features of neonatal abstinence syndrome did not occur. The woman remains symptom-free three months postpartum while taking prednisone 5 mg daily.

Figure 1.

Interstitial glucose monitoring demonstrating hypoglycaemia.

Discussion

Hypoglycaemia related to non-diabetic pregnancy is rare, the majority of reported cases due to insulinoma, acute fatty liver of pregnancy, malaria and inborn errors of metabolism. HPAi may also occur postpartum due to lymphocytic hypophysitis or Sheehan’s syndrome. Maternal hypoglycaemia in animal studies is associated with increased risk of congenital malformations though it is unclear whether there is any increased risk in human pregnancy. ¹ Maternal hypoglycaemia has been associated with fetal growth restriction and small for gestational age infants.

The incidence of opioid dependence in pregnancy is increasing dramatically. ² Reproductive-aged women experienced a 400% increase in overdoses from prescription opioid medications between 1999 and 2010, and the number of pregnant women reporting opioid misuse has risen 130% over the last two decades.^2,3 Pregnant women may not disclose opioid use because of social stigma, potential legal consequences, and child custody issues. The Society for Obstetrics and Gynecology of Canada and the American College of Obstetrics and Gynecology endorse universal screening for substance use in pregnancy. Opioid dependence is associated with pre-term birth, premature rupture of membranes, placental abruption, preeclampsia, small for gestational age infants, reduced head circumference, fetal death, neonatal abstinence syndrome and sudden infant death syndrome.

Mechanisms of opioid-associated hypoglycaemia include HPAi, reduced hepatic glycogen, impairment of counter-regulatory responses, and hyperinsulinaemic, hypoketotic hypoglycaemia with elevated C-peptide has been described with methadone, tramadol and propoxyphene.^4–6 The latter mechanism may therefore manifest a similar hormonal profile as seen with dumping syndrome, insulinoma, nesidioblastosis or surreptitious sulphonylurea ingestion. HPAi occurs in 9–29% of individuals with chronic opioid use. ⁷

Venous glucose below 2.6 mmol/L occurs in 10% of healthy non-pregnant individuals following glucose challenge . ⁸ The prevalence of hypoglycaemia following bariatric surgery is unclear due to variability in definitions of hypoglycaemia (level of glucose, symptoms), variance in surgical procedures, and whether hypoglycaemia was precipitated by a glucose challenge or standard diet. Symptomatic hypoglycaemia following bariatric surgery in non-pregnant individuals consuming a normal diet is uncommon, having been reported in only 0.1–1.2%, and is far less common following LSG than gastric bypass surgery.^9–11 While hypoglycaemia following glucose challenge has been reported in 83% of pregnant women with previous gastric bypass surgery and 55% of women with previous LSG, the prevalence of symptomatic hypoglycaemia in pregnancy following bariatric surgery while consuming a normal diet is unknown, having only been described in a few case reports .^12–14 The presence of reactive hypoglycaemia should therefore be sought with a mixed meal and not a glucose challenge.

A physiological rise in serum cortisol and cortisol-binding globulin of approximately 3-fold occurs during pregnancy. Trimester-specific reference intervals for serum cortisol with spot early morning measures and post-synacthen stimulation must be employed to avoid missing diagnosis of HPAi in pregnancy, and if concerns remain, empirical treatment until confirmatory tests can be performed after delivery. No studies have been performed regarding the requirement for changes in replacement doses in pregnancy, although Endocrine Society guidelines recommend glucocorticoid doses may need to be increased by 20–40% in third trimester. ¹⁵ Salivary cortisol day curves have been used to assess glucocorticoid replacement in Addison’s disease in non-pregnant subjects. ¹⁶ Several studies have examined salivary cortisol levels throughout healthy pregnancy, suggesting measures may provide a cortisol-binding globulin independent measure of adrenal function. ¹⁷ Salivary cortisol day curve may have been useful in this case to assess the adequacy of the woman’s hydrocortisone replacement.

Systemic corticosteroid use during pregnancy has been linked with increased risk of preterm birth; however, it is unclear whether this is related to the underlying medical condition for which corticosteroids were prescribed or the medication itself. ¹⁸ Preterm birth has been reported in 14% of women treated with buprenorphine, compared with 21–25% of methadone exposed pregnancies.^19,20

In conclusion, a case of recurrent non-diabetic hypoglycaemia in the setting of previous bariatric surgery and opioid-dependence which resolved symptomatically following the use of a longer-acting corticosteroid in a proportionally higher dose is described here. It is possible that the hypoglycaemia in the woman presented may have been due to an alternative mechanism; however, the absence of symptoms having weaned to a near-physiological replacement dose of prednisone would make insulinoma and non-islet-cell tumour hypoglycaemia unlikely. Health professionals should be aware of the increasing prevalence of opioid misuse during pregnancy and the possibility of non-disclosure. Opioid use may cause hypoglycaemia by multiple mechanisms associated with varying biochemical and hormonal results. Glucose challenge should not be used to diagnose hypoglycaemia due to late dumping following bariatric surgery, a rarely reported cause of hypoglycaemia in pregnant women. Trimester-specific serum cortisol values must be used to exclude HPAi in pregnancy. Guidelines regarding requirements for changes in dose and/or dose frequency of glucocorticoid replacement in pregnant women with HPAi do not appear to have been based upon scientific studies. Future research assessing the usefulness of salivary cortisol day curves in adjusting glucocorticoid replacement in pregnant women with HPAi would be valuable.